Posts: 9
Joined: Thu Jan 28, 2021 4:16 am

Understanding battery capacity based on SOC and voltage

Tue Feb 09, 2021 12:28 pm

Hi all,

Using the PHEV Watchdog I recorded some voltage measurements for a range of reported SOC values. The measurements were taken after an at least 3h rest after charging or driving, so that the battery can rest and the BMU has time to reevaluate the SOC. The range goes from 20% SOC (the lowest I was able to get) up to full charge. As you see in the attached plot, in the studied range the points are nicely fitted with a linear function (blue line). Moreover, SOC at 4.1V is 31.1Ah, which is exactly what the dog reports as my battery capacity (77.8%).

The red line has the same slope, but is shifted so that capacity at 4.1V is 40Ah instead of 31.1Ah. The point of this: IMO if we could add data obtained from two more cars (1) a new car, and (2) a car after the manual Lindqvist capacity reset, we could figure out how well the BMU is doing in estimating the true battery capacity. This could also provide us with means to build an independent battery capacity model.

If I would subject my car to the Lindqvist reset, the new measurements would plot on the shifted red line in the figure, give or take the collapse of the linear fit in the left side of the plot for low SOC (the SOC-Voltage relationship is no longer linear for low SOC). This would essentially mean that the reset simply causes a deeper discharge of the lower buffer. If we had measurements for such a car, and we could actually observe the non-linear regime, it would give us insight into the true SOC, and consequently true SOH.

On the other hand, a new car should show a different fit. Assuming that the BMU attempts to keep the voltage at the bottom of the discharge the same for all cars (~3.8V) and that obviously the same voltage is put as the roof when charging (4.1V), new cars should plot roughly on the black dashed line in the second figure. Another option would be that the BMU tries to keep a constant charge as the lower buffer (here around 6.3Ah). This would mean that the voltage at the bottom would be have to vary. Again, having such data for a new car would tell us what it is that the BMU is doing.

Would anyone be interested to contribute some data points? Before you record the data, please wait at least 3h after charging or driving.
You do not have the required permissions to view the files attached to this post.

Posts: 1
Joined: Mon Feb 15, 2021 5:39 am

Re: Understanding battery capacity based on SOC and voltage

Tue Feb 16, 2021 6:23 am

I'd be more than happy to contribute. I just got a MY17 and want to start taking some measurements anyway to see how healthy the battery is.

Posts: 135
Joined: Sat Oct 27, 2018 3:47 am

Re: Understanding battery capacity based on SOC and voltage

Sat Jun 05, 2021 2:34 am

Outlander serviced yesterday they supplied me with a battery test slip.

Hybrid. 12v Aux Test
Voltage 12.33
Measured 363 CCA

Rate S46B24
Temperature 22 C

Battery Location In Vehicle
Battery type AGM

Test Code O3U94 - J1)J0-CGGK8H
State of health good
SOC 75%

I have charged up this morning 10.2 kWh to charge, I regularly achieve around 10.8 kWh.
Mitsubishi Outlander 2.4 PHEV Design 2020 Towing a Swift Conqueror 480 2016 1500kg

Please be patient I suffered a brain injury several years ago and get confused at times

Posts: 606
Joined: Wed Sep 09, 2015 5:05 pm
Location: Australia

Re: Understanding battery capacity based on SOC and voltage

Sat Jun 05, 2021 2:52 pm

I'm pretty sure that is the 12V battery, not your 360V drive battery.

I've noticed a 'battery health' statement when I get mine serviced, but it is not relating to the drive battery.
(I've asked them)
2014 PHEV Aspire, Arctic Silver;
Black Interior;
Hayman Reese removable tow bar;
1.5kW Solar, grid connected.

Return to “Technical Discussions”